1 Introduction

The melt blowing process is a commercial single-step process that fabricates polymer nonwoven products with superfine fibers with a diameter <5 μm (and even below 1 μm). The polymer is melted and extruded by the extruder and then is drawn into the hot, high-speed air jet. Suction fans under the grid-like collection belt assist adhesion, by which the fibrous web “locks” on the collection belt. The web is then strengthened by bonding with two hot rollers, generating the final melt-blown nonwoven material. In the melt blown industry, basis weight distribution (BWD) or variation coefficient of basis weight (BWCV) for the fibrous web is always used to evaluate the uniformity of fibrous webs. The uniformity of the fibrous web largely determines the appearance and the functional properties of the final MB products. For instance, in the filtration process, a non-uniform filter may lose its efficacy due to allowing large particles to pass through the thin spots in the fiber sheet.

In recent years, the uniformity of fibrous webs has been widely studied using image analysis techniques in many papers. For instance, Bresee (1) and Cherkassky (2) studied the basis weight and other web structural features, including pore size distribution, fiber orientation angle, fiber diameter and so on, based on experiments and image analysis. Similarly, Chabbra (3) proposed two parameters called the standardized index of dispersion and the mass anisotropy ratio to study the BWD of fibrous webs using image analysis technique based on reflected/transmitted light of fibrous webs. Two new algorithms called learning vector quantization and the fuzzy neural network were also employed during analyzing the images of nonwoven materials in the work of Liu and co-workers (4), (5) etc.

Most researchers focus on the mathematical algorithms. However, image analysis concerning solving engineering problems such as the influence of processing conditions on the uniformity of the melt-blown fibrous webs is seldom reported. Therefore, in this study, the uniformity of MB fibrous webs collected under different processing conditions is quantitatively studied based on an image analysis technique. The optimal processing conditions could be obtained according to the analysis results.

2 Experiment

An experiment was designed for capturing upward view images of fibrous webs collected under different processing conditions. We prepared the fibrous web using a melt-blowing machine with a single-orifice dual-slots die (F-6D, Huada Co., Ltd., Yantai, China). The processing conditions are shown as follows: nominal air jet pressure with the range of 0.15 MPa–0.4 MPa, volume flow rate of polymer with the range of 0.065 ml/min–0.39 ml/min, air suction velocity with the range of 1 m/s–3 m/s and moving speed of collection belt with the range of 8 cm/s–12 cm/s. Moreover, the collector was stationary except for the experiment performed under the last processing condition. Then, the fibrous web was moved onto a glass screen in a dark box after collection. There was a 20-watt lamp above the dark box and a digital camera (ILCE-600, SONY, Tokyo, Japan) was under the glass screen, through which upward view photographs of the fibrous web were obtained. A sketch of the image capture setup is shown in Figure 1A. The obtained resultant image of the fibrous web is shown in Figure 1B. In addition, in order to verify our image analysis method, four fibrous webs were cut into 4×4 pieces. Each piece was weighed according to standard ISO 9073-1. The results of verification experiment are given in the Section 4.1.

Figure 1:

Image capturing and processing: (A) photographing experiment; (B) image processing.

3 Approach of image analysis

The gray level of the central area of the original image of Figure 1B was smaller than that of the circumambient area. But the fiber density of the central area was larger than that of the circumambient area. The variation of fiber density was contrary to the variation of gray level. Therefore, a revised image was generated using 255 minus each gray level of original image. A white background (the gray level is larger than 200) was considered as 0 during calculation. The parameters of BWCV of each revised image were calculated based on the following equations.

Where

Vm=∑i=1m∑j=1nVij4

stdCD=∑i=1m(∑j=1nVij-Vm)2

stdMD=∑j=1n(∑i=1mVij-Vm)2

Herein, CD represents cross direction, MD represents machine direction. i and j represents the pixel position of image, respectively. The image is a matrix with m×n size and V is the gray level of image. std is the standard deviation of the gray level. Smaller BWCV represents a more uniform fibrous web, whereas a non-uniform fibrous web always has a larger BWCV. Calculated BWCV_CD and BWCV_MD could be used to describe the difference of the uniformity along CD and MD of fibrous webs.

4 Results and discussion

4.2 The influence of process conditions on the uniformity of fibrous webs

4.2.1 The influence of nominal air jet pressure

Figure 2 shows the BWCV of fibrous webs collected under different nominal air jet pressure. It can be seen from Figure 2 (the experimental data is shown in Table S5 of part C in the Supplementary material) that BWCV was decreasing rapidly with increasing nominal air jet pressure. That means a worse web along both the slot and machine direction is generated under a smaller nominal air jet pressure. On the contrary, a more uniform web was formed under larger nominal air jet pressure. This is reasonable as the evenly distributed pores and fine fibers of the fibrous web result in a more uniform fibrous web with smaller BWCV values. According to the work of Bresee (6), (7), (8), (9), small and evenly distributed pores and fine fibers were observed in the fibrous web collected under larger nominal air jet pressure. The calculated data demonstrated that BWCV_MD was smaller than BWCV_CD. Because continuous filaments are filled with the non-uniform part of the fibrous web on the moving collector in the MB process, the non-uniform part of fibrous web along MD is improved, which leads to smaller BWCV_MD. It is both in line with our observation and the study of Chahabra (2).

Figure 2:

The influence of nominal air jet pressure.

4.2.2 The influence of polymer volume flow rate

Figure 3 illustrated that the swelling of the polymer volume flow rate resulted in the steady increase of BWCV. The experimental data is shown in Table S6 of part C in the Supplementary material. In other words, the uniformity of the web structure is worse when the polymer volume flow rate is increasing. On the basis of the experiment of Rao (10), the increase of the polymer volume flow rate leads to the fiber diameter going up. Overlap of the thick fiber brings about a coarse pore structure, which generates a non-uniform fibrous web with a large BWCV value. Calculated BWCVs also exhibit larger BWCV_CD but smaller BWCV_MD.

Figure 3:

The influence of polymer volume flow rate.

4.2.3 The influence of movement of the collection belt

As shown in Figure 4, the BWCV (the original data is listed in Table S7 in the Supplementary material) decreased with the increase of the speed of the collection belt, revealing the uniformity of the fibrous web was improved when the speed of collection belt was increasing. It is because the fiber tends to land on the central part of the collection belt. This phenomenon is more obvious when the collection belt is slower. Therefore, the fibrous web appears to be nonuniform. However, the uniformity is improved when the collection belt is faster. Additionally, a more uniform fibrous web (smaller BWCV value) is also generated on the collection belt with higher speed, especially along the MD.

Figure 4:

The influence of movement of collection belt.

4.2.4 The influence of air suction under the collection belt

As shown in Figure 5 (detailed data is given in Table S8 in the Supplementary material), BWCV was declining when the speed of air suction was growing, which revealed that a more uniform fibrous web was generated. Many fibers are fixed on the collection belt, while a few fibers are flown into the surrounding air due to the high speed of air suction under the collection belt. Therefore, the uniformity of the fibrous web is improved with high speed air suction. But the change of BWCV is limited at higher air suction due to the declining slope of the fitted curve. In addition, the calculated BWCVs also suggest that the uniformity of fibrous web in CD is worse than that in MD.

Figure 5:

The influence of air suction.

5 Conclusion

In this study, the influence of processing conditions on the uniformity of the MB fibrous web is studied based on the image analysis technique, including nominal air jet pressure, polymer volume flow rate, moving speed of collection belt and air suction velocity. The results suggest that with the soaring of air jet pressure, the basis weight of fibrous web is distributed more evenly. The same situation is observed under the condition of the speed of collection belt and velocity of air suction. On the other hand, the uniformity of the fibrous web is worse when the polymer volume flow rate is increasing. According to the results, it seems that higher air jet speed, higher air suction velocity, lower polymer volume flow rate and faster moving collection belt could help improve the uniformity of fibrous webs. Additionally, under the studied process conditions, the fibrous web along the machine direction is more uniform than that of the cross direction. Our analytical work could give a better reference to the commercial MB process.